January

2015

Genomic Selection Drives Martin Dairys

Milk Production for Tillamook Cheese
Martin Dairy uses Infinium BovineSNP50 BeadChip technology to improve breeding decisions
and build one of the most productive Jersey herds in the United States.
Introduction
Norman Martin comes from a long line of dairy farmers who
made their way from Portugal to California, finally settling in
the Pacific Northwest. His farm is on land purchased in
Tillamook, Oregon in 1995 to raise and breed Holstein and
Jersey cattle. Eventually converting to a 100% Jersey herd
in 1998, the herd has grown from 80 to about 2200 cows.
Today, Martin Dairy
is one of about 100 milk providers to the farmer-owned
dairy co-operative, Tillamook County Creamery
Association1, home of world renowned Tillamook cheese.
Working hand-in-hand with the American Jersey Cattle
Association2, Martin Dairy was an early adopter of genomic
technologies to improve dairy selection decisions. Norman
first used the BovineSNP50 BeadChip to genotype two
cows in March 2009. To stay ahead of the curve and expand
its use of the technology for dairy cattle breeding, Martin
Dairy continued incorporating new technology and currently
uses low-density arrays based on the original BovineSNP50
BeadChip. To date, the dairy has accumulated data on over
90% of its herd. Moreover, the Martin Dairy Jersey herd is
the fourth highest genetically ranked herd in the United
States, and sells seedstock, semen, and embryos
internationally
for artificial insemination (AI). Norman attributes his farms
success to the efficiency of the Jersey breed, and the
advantages genomic selection provides in accelerating the
genetic progress of his herd above breed averages for the
economically important traits that are accelerating popularity
of the breed globally.
iCommunity spoke with Norman to learn about the
intricacies of the Jersey breed and how he has used
genomics to maximize profits through increased productivity.
Q: Whats the history behind Martin Dairy?
Norman Martin (NM): We moved from California to Oregon
in 1995 when we bought 160 acres, but weve added 140
since then. I only brought up 80 cows. At the time, I thought
that milking a herd that was half Holsteins and half Jerseys
would be the right thing to do, given the milk market here.
But after a couple of years I found that I was wrong. Now
we are 100% Jersey.

Norman Martin (right) is owner of Martin Dairy in Tillamook, Oregon.

He and his son Chad manage a herd of more than 2000 Jersey
cows.

Q: How did you begin providing milk to Tillamook Cheese?

NM: Weve provided milk to Tillamook since we moved the
farm up here. All you have to do to become a member of
the co-op is own a farm in Tillamook County and then
become approved by the board. Theres a lot of history that
goes
into Tillamook. At one time, there were 27 cheese plants
in Tillamook County and 800 dairy farms. Now, everybody
in the County ships their milk to the co-op, so all the dairy
farmers are the owners of Tillamook Cheese. We have
one
plant and about 100 members. Becoming a member is easy.
If you dont meet the quality standards, youll be out of the
co-op just as quickly.
Q: What was it about Jersey cows that made you switch
to a 100% Jersey operation?
NM: I kept hearing a buzz about Jerseys while I served
as chair of the Holstein Sire Committee for probably
about a dozen years. Id always sit in on the Jersey
meetings after mine because I was curious.
Fundamentally, it came down to yield and efficiency.
Approximately 80% of all the milk produced in the United
States goes into manufacturingcheese, butter, yogurt,
and the like. In order to manufacture these products, you
need a high yield of the right milk components.

For Tillamook Cheese, our vats are 50,000 pounds. If you fill
them with Jersey versus Holstein milk, you end up with 1000
pounds more cheese. When youre processing 28 vats a day,
thats 28,000 pounds more cheese at the end of 24 hours for
the same amount of time and money. Because they offer a
higher yield of the right components, Jersey cows enable us
to have a more efficient operation.

Because they offer a higher

yield of the right components,
Jersey cows enable us to have
a more efficient operation.
Q: When did you begin using genomic selection tools in
your breeding program?
NM: Whenever new technologies come along, I usually get
right on them within a short period. Its part my personality,
and part surviving in the business world. If you want to stay
ahead of the game, you better get on the train quickly.
We did some of the initial genotyping work in late 2007
using some of the first tools, such as microsatellite
analysis. Then we took the whole-genome approach and
did preliminary testing with the 3K chip. Our first official
genomic evaluations for Jerseys were in January 2009 and
the first 2 females were genotyped from hair samples in
March 2009 using the
BovineSNP50 BeadChip. Their first genomic evaluations
were produced in April 2009. These 2 cows now have more
than 50 combined progeny, 9 sons in AI, and more than
1300 combined grandchildren.
We currently use low-density array systems with content
based on the original BovineSNP50 BeadChip. To date,
1991 Jersey males have been genotyped by Martin Dairy,
including 85 fourth generation- and 13 fifth generation-bulls.
Ninety-four of our bulls have gone to national studs and
entered into AI service, with over 10,455 registered Jerseys
being sired by
a bull originating from our herd.

We currently use lowdensity array systems with

content based on the
original BovineSNP50
BeadChip.
Q: How do genomics tools integrate with the new
reproductive technologies that you are using to enhance
the genetic value of your herd?

NM: We have been using embryo transfer in our cows with the
highest genetic value for several years, so we have higher
genomic level calves coming through. We genotype all of

our heifers and use a genetic merit measurement called

a
P.L evel to select which will enter our sex-sorted AI breeding.
We breed about 60 heifers each month, with those
having
a P > 7 (on a 09 scale) bred with sex-sorted semen.
These heifers rank in the top 20% of the Jersey breed.
Breeding with sex-sorted semen means that we try to
breed more females out of the high-level heifers,
creating cows with even higher genetic values. With
genotyping, were now able to identify young sires that
have higher genetic levels and use those in our next
generation with even more confidence
than before.
Q: What traits are you selecting for with genomics tools?
NM: Money is the main trait. I select for cows that will
wear well and produce many pounds of liquid cheese
that I can ship to the plant. We have many cows that
stay around here until theyre 8 or 9 years of age and
then all of a sudden they get
to where they cant compete anymore.
Also, feed efficiency is important to everybody in the
dairy business. Feed is our biggest operating cost, and
the Jersey breed is the smallest breed. So if you want
to look at it as food per pound of cow versus how
many pounds of cheese produced, Jerseys are
number one in the world.

With genotyping, were now

able to identify young sires

that have higher genetic levels

and use those in our next
generation with even more
confidence
than before.
Q: Who does the data analysis and how is it integrated
with the phenotype information?
NM: The American Jersey Cattle Association, which Im so
proud to be a part of, provides us with all these tools. Ill
go through the different AI studs when new bulls come
through. I look at their genotypes and pedigrees, make
decisions on which ones I want to use, and then Ill send a
list of heifers that Im breeding every month to the Jersey
Association and the 15 or 20 bulls that I plan on using.
They send me a list
back with three different choices, and I make the final
decision among the three. Because I run two different
groups at a time, I end up with six choices.
Its the genotype that gives me all the information to make
breeding decisions and manage inbreeding, which is
always an issue. My system gives me all these choices. If
I use bull
#1, I get this, if I use bull #2 I get that, and if I use bull #3 I
get something else entirely, and I make decisions from
there. The resulting genetic evaluation uses phenotypic
data along with

January 2015
the genomic information. So its an integrated product in
the end when you get a breeding value.
Q: What is the value for you in comparing your herd
with other herds or the entire US Jersey population?
NM: Like anything else, everybody wants the gold
medal.
Ive always been that way. If someone has a car that can go
200 mph, I want mine to be able to do 202 mph. Thats what
its like in the dairy businessproduction and stayability of
cows is where money comes from and thats the reason
were in business.

Now, because of the genetic

information that I have, my son
Chad and I can confidently
determine that a highly rated
genomic young sire is
equivalent to a proven sire and
we dont have to wait.
Q: As a past board member of the Jersey Association,
can you talk about the Associations leadership role in the
implementation of genomics to Jersey genetic evaluation?
NM: The Association has always been on the cutting
edge.
But if youd asked me some of these same questions 40
years ago, Id have told you that the Jersey breed would
likely not be here now. Best I can tell, its probably without a
doubt the fastest growing dairy breed in the world.
A lot of people in the right places have made good
decisions about where the breed should go. Weve seen
other breeds, for instance Guernsey dairy cattle, go from
being one of the leading breeds in the United States in the
1940s to now being on the watch list of The Livestock
Conservancy3.
Q: How much is the genomic data on these young bulls
driving the process of selling them into AI now?
NM: Genomic data influences AI decisions more every day. I
really believe that. Theres always some hesitation when
these new things come along, but Im full bore into it. When
most people buy their semen, they are looking at the
genomic level of that bull. Everybody has different ways of
mating cows; some are more concerned about udders while
some might be concerned about feet and legs so selection
processes differ as well. About 95% of the bulls that are even
being considered for sampling at this point are genotyped.
So its a major driver.

In the early days of genomic testing, I would only use sires

that were proven sires based on the production
performance of their offspring. That was true even though
the genetics were giving me a clear indication of their value.
Now, because of
the genetic information that I have, my son Chad and I can
confidently determine that a highly rated genomic young
sire is equivalent to a proven sire and we dont have to
wait.
Q: How do you keep track of the mating pairs and progeny?
NM: Every day, every cow that is bred or served on that
day gets entered into the computer. If someone from an
AI organization asked me right now how many cows I
have
confirmed pregnant to a specific bull, within about 15
seconds I can tell them exactly how many. How many of
those are carrying heifers? Again, same answer. The system
I use is Dairy Comp 3054.
The Jersey Association that maintains all this data and
their services are outstanding. Im always waiting to see
whats coming up next. We used to wait a year to get
evaluations on cattle, then we got to where we evaluated
them every six
months, then it was every quarter. When we got down to
once a month, I thought wow, it isnt going to get any better
than this. Yet, now were performing evaluations once a
week. I dont think we need to go to performing them daily,
but I could be wrong.
Q: What types of reports do you provide when you sell bulls
into AI?
NM: We provide the pedigree and everything else there is to
know about the animal, and then they run their own
genomic tests. It does make us a little money, but its not a
driving force behind our business. Were a commercial dairy
operation selling milk to a high profile co-op and cheese
label, and thats the first thing we need to be concerned
aboutproviding a high-quality product to our creamery.

It doesnt take much for a

mechanic to get a car to go from
45 mph up to 70 mph, but it
takes a fine-tuned, savvy
individual to get a car to go from
137 to 141 mph and win the Indy
500. The same is true for
genomics and cows.

January 2015
Q: How do you see genomics playing into the future at
Martin Dairy?
NM: Well continue to do what were doingpairing the very
best with the very best, over and over. It doesnt take much
for a mechanic to get a car to go from 45 mph up to 70 mph,
but it takes a fine-tuned, savvy individual to get a car to go
from 137 to 141 mph and win the Indy 500. The same is true
for genomics and cows. As we move farther up the ladder
and our genomic scores become higher, our amount of gain
each time will be slightly less. So you have to strive a little
harder to make each improvement.
I dont see anything negative in shortening generational
intervals and speeding up genetic progress. But there is no
ultimate goal, because there is always the next step to
take.
It never ends. The selection of dairy cattle has been going
on as long as there have been dairy cows on Earth, and
the faster it goes, the more information you have to make
it go even faster.
Q: What would be your advice to young people today
wanting to enter the dairy business?
NM: The most important thing is to be a commercial
dairyman first and make your money selling milk. Next, use
the best bulls you can afford to use and continue to do that
over and over.
Thats what Ive been doing basically all my life.

references
1.

Tillamook Dairy, www.tillamook.com/

2.

American Jersey Cattle Association, www.usjersey.com/

3.

The Livestock Conservancy cattle conservation priority list,

www.livestockconservancy.org/index.php/heritage/internal/conservationpriority-list#Cattle.

4.

VAS DairyComp 305, www.vas.com/dairycomp.jsp

learn more about the Illumina product mentioned

in this article:
Infinium BovineSNP50 BeadChip, www.illumina.com/
products/bovine_snp50_whole-genome_genotyping_kits.
html

Follow the Illumina Agrigenomics Community

Illumina 1.800.809.4566 toll-free (U.S.) +1.858.202.4566 tel techsupport@illumina.com www.illumina.com

For research Use oNly. NoT For IN VITro DIaGNosTIc Use.
2015 Illumina, Inc. All rights reserved.
Illumina, Infinium, and the pumpkin orange color are trademarks of Illumina, Inc. and/or its affiliate(s) in the U.S. and/or other
countries. All other names, logos, and other trademarks are the property of their respective owners.
Pub. No. 1370-2014-015 Current as of 01 April 2015